CN118225254B - Multi-parameter anomaly calibration system and method applied to thermal infrared imager - Google Patents
Multi-parameter anomaly calibration system and method applied to thermal infrared imager Download PDFInfo
- Publication number
- CN118225254B CN118225254B CN202410552357.4A CN202410552357A CN118225254B CN 118225254 B CN118225254 B CN 118225254B CN 202410552357 A CN202410552357 A CN 202410552357A CN 118225254 B CN118225254 B CN 118225254B
- Authority
- CN
- China
- Prior art keywords
- control behavior
- behavior information
- information
- focal plane
- classification
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000011156 evaluation Methods 0.000 claims abstract description 101
- 238000012216 screening Methods 0.000 claims abstract description 16
- 238000013461 design Methods 0.000 claims abstract description 14
- 230000000694 effects Effects 0.000 claims description 37
- 238000003384 imaging method Methods 0.000 claims description 33
- 230000008569 process Effects 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000033228 biological regulation Effects 0.000 abstract description 7
- 230000004069 differentiation Effects 0.000 abstract description 3
- 230000002159 abnormal effect Effects 0.000 abstract description 2
- 230000006399 behavior Effects 0.000 description 146
- 230000009471 action Effects 0.000 description 14
- 241001394244 Planea Species 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J5/00—Radiation pyrometry, e.g. infrared or optical thermometry
- G01J5/90—Testing, inspecting or checking operation of radiation pyrometers
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Radiation Pyrometers (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a multi-parameter anomaly calibration system and method applied to a thermal infrared imager, and belongs to the technical field of thermal infrared imagers. Based on the internal structural design of the thermal infrared imager, generating a position label and an angle label of a focal plane, and recording control behavior information of a user side; randomizing the control behavior information into a plurality of groups, wherein each group forms a sub-set data cluster, analyzing the data quality, and carrying out quality screening on the control behavior information; classifying the control behavior information based on the screening result; evaluating the control behavior information through an evaluation index, and calibrating the control behavior based on an evaluation result; therefore, the method can make up the limitation of hardware, realize self-adaptive optimal regulation and control on software, and can realize the maximum range of unified index parameters based on the unified hardware limitation, so that the differentiation of the combination regulation and control of the software and the hardware is reduced as much as possible, and the abnormal calibration of the thermal infrared imager is realized.
Description
Technical Field
The invention relates to the technical field of thermal infrared imagers, in particular to a multi-parameter anomaly calibration system and method applied to the thermal infrared imagers.
Background
The infrared thermal imager realizes an imaging function through a focal plane array, and optimizes the induction behaviors of each detector unit in the focal plane array through software, so that focal plane compensation or calibration can be realized;
Then, the focal plane hardware is limited, so that the compensation or calibration on the software is difficult to optimize based on the unified index parameters, and in many cases, the maximum limit optimization is realized on the software, but the selection of various control modes still exists on the hardware.
Disclosure of Invention
The invention aims to provide a multi-parameter anomaly calibration system and a multi-parameter anomaly calibration method applied to a thermal infrared imager so as to solve the problems in the background art.
In order to solve the technical problems, the invention provides the following technical scheme:
A multi-parameter anomaly calibration system applied to a thermal infrared imager comprises: the system comprises an operation log module, a data quality analysis module, a data processing module and a parameter calibration module;
The operation log module is used for generating a position label and an angle label of a focal plane based on the internal structural design of the thermal infrared imager and recording control behavior information of a user side;
the data quality analysis module is used for randomizing the control behavior information into a plurality of groups, forming a subset data cluster by each group, analyzing the data quality and carrying out quality screening on the control behavior information;
The data processing module classifies the control behavior information based on the screening result;
the parameter calibration module is used for evaluating the control behavior information through the evaluation index and calibrating the control behavior based on the evaluation result.
Further, the operation log module comprises a cache library unit and a tag pair unit;
The cache library unit is used for establishing an operation log cache library, wherein the operation log cache library stores control behavior information during each dynamic adjustment of imaging effect, and the control behavior information comprises position information of a focal plane and angle information of the focal plane; based on the internal structural design of the thermal infrared imager, respectively establishing a position tag and an angle tag of a focal plane in a unified form, wherein the position tag and the angle tag respectively have unique tag attributes;
The label pair unit is used for marking the ith position label as The j-th angle label is marked asCollecting control behavior information during the kth dynamic adjustment of imaging effect, generating a control behavior set, and recording asAnd (2) andWherein, the method comprises the steps of, wherein,E represents the E-th control behavior information at the kth time of dynamically adjusting the imaging effect, E represents the total number of control behavior information at the kth time of dynamically adjusting the imaging effect,Position tag representing focal planeAngle label for sum focal planeAnd (5) forming a label pair.
Further, the data quality analysis module comprises a subset construction unit and a quality analysis unit;
The subset forming unit is used for constructing a behavior representation two-dimensional coordinate system based on control behavior information, wherein the abscissa of the behavior representation two-dimensional coordinate system corresponds to a position label of a focal plane, and the ordinate of the behavior representation two-dimensional coordinate system corresponds to an angle label of the focal plane; marking control behavior information in the behavior representation two-dimensional coordinate system based on the behavior representation two-dimensional coordinate system; for control action set Each control behavior information in the data cluster is randomly grouped to form a subset data cluster, and any subset data cluster is obtained and recorded asAnd (2) and，X represents the number of the subset data clusters, and y represents the total number of the subset data clusters; to control behavior informationAs a subset data clusterRespectively obtain subset data clustersPunctuation corresponding to each control behavior information and control behavior informationEuclidean distance between corresponding punctuations, and based on the Euclidean distance, obtaining a subset data clusterIs recorded as the intra-cluster distance of (2)；
The quality analysis unit is used for constructing a data quality iterative analysis model, and controlling a behavior set during each iterationA sample input is selected at will, and the sample input selected in each iteration is different;
Control behavior information As a sample input for the r-th iteration, a data quality value is calculated, and a specific calculation formula is as follows:
wherein, Representing control behavior informationIs used for the data quality value of (a),A desired value representing a minimum intra-cluster distance;
presetting a data quality threshold, and extracting control behavior information if the data quality value is greater than or equal to the data quality threshold And stores the output information setIn the process, otherwise, the control behavior information is removed；
Up to the control action setAfter each piece of control behavior information participates in iteration, the iteration is stopped.
Further, the data processing module comprises a first classifier unit and a second classifier unit;
The first classifier unit is used for outputting information sets Performing first behavior classification, taking the position information of the same focal plane as a first classification reference, traversing label pairs with the same first classification reference, generating a first classification set, and recording asWherein, the method comprises the steps of, wherein,Representing position information in focal planeCorrespondingly generating a first classification set for the first classification reference;
the second classifier unit is used for outputting information sets Performing second behavior classification, traversing tag pairs with the same second classification references by taking angle information of the same focal plane as the second classification references, generating a second classification set, and recording asWherein, the method comprises the steps of, wherein,Representing angular information in the focal planeFor the second class reference, the generated second class set corresponds.
Further, the parameter calibration module comprises an index parameter unit and a calibration analysis unit;
the index parameter unit is used for establishing an evaluation index parameter category information base, and index parameter category information for evaluating the control effect is stored in the evaluation index parameter category information base; uniformly coding the evaluation index parameters to form an evaluation index parameter set with unique coding attribute, and recording the evaluation index parameter set as Wherein, the method comprises the steps of, wherein,Represents the a-th evaluation index parameter, A represents the total number of evaluation index parameters;
The calibration analysis unit is used for evaluating the index parameter set Each of the evaluation index parameter pairs controls behavior informationEvaluating to obtain control behavior informationIs a mean value of the initial evaluation of (a); obtaining position information based on initial evaluation mean value of each control behavior information in the first classification setIs recorded as the comprehensive evaluation mean value of (2); Obtaining angle information based on the initial evaluation mean value of each piece of control behavior information in the second classification setIs recorded as the comprehensive evaluation mean value of (2)；
Respectively byAndAnd calibrating the positions and angles of the focusing planes of the corresponding position labels and the angle labels.
The multi-parameter anomaly calibration method applied to the thermal infrared imager comprises the following steps of:
Step S100: based on the internal structural design of the thermal infrared imager, generating a position label and an angle label of a focal plane, and recording control behavior information of a user side;
Step S200: randomizing the control behavior information into a plurality of groups, wherein each group forms a sub-set data cluster, analyzing the data quality, and carrying out quality screening on the control behavior information;
step S300: classifying the control behavior information based on the screening result;
Step S400: and evaluating the control behavior information through the evaluation index, and calibrating the control behavior based on the evaluation result.
Further, the specific implementation process of the step S100 includes:
Step S101: establishing an operation log cache, wherein the operation log cache stores control behavior information during each dynamic adjustment of imaging effect, and the control behavior information comprises position information of a focal plane and angle information of the focal plane; based on the internal structural design of the thermal infrared imager, respectively establishing a position tag and an angle tag of a focal plane in a unified form, wherein the position tag and the angle tag respectively have unique tag attributes;
Step S102: marking the ith position tag as The j-th angle label is marked asCollecting control behavior information during the kth dynamic adjustment of imaging effect, generating a control behavior set, and recording asAnd (2) andWherein, the method comprises the steps of, wherein,E represents the E-th control behavior information at the kth time of dynamically adjusting the imaging effect, E represents the total number of control behavior information at the kth time of dynamically adjusting the imaging effect,Position tag representing focal planeAngle label for sum focal planeAnd (5) forming a label pair.
Further, the specific implementation process of the step S200 includes:
Step S201: constructing a behavior representation two-dimensional coordinate system based on control behavior information, wherein the abscissa of the behavior representation two-dimensional coordinate system corresponds to a position label of a focal plane, and the ordinate of the behavior representation two-dimensional coordinate system corresponds to an angle label of the focal plane; marking control behavior information in the behavior representation two-dimensional coordinate system based on the behavior representation two-dimensional coordinate system; for control action set Each control behavior information in the data cluster is randomly grouped to form a subset data cluster, and any subset data cluster is obtained and recorded asAnd (2) and，X represents the number of the subset data clusters, and y represents the total number of the subset data clusters; to control behavior informationAs a subset data clusterRespectively obtain subset data clustersPunctuation corresponding to each control behavior information and control behavior informationEuclidean distance between corresponding punctuations, and based on the Euclidean distance, obtaining a subset data clusterIs recorded as the intra-cluster distance of (2)；
Step S202: constructing a data quality iterative analysis model, and controlling a behavior set during each iterationA sample input is selected at will, and the sample input selected in each iteration is different;
Control behavior information As a sample input for the r-th iteration, a data quality value is calculated, and a specific calculation formula is as follows:
wherein, Representing control behavior informationIs used for the data quality value of (a),A desired value representing a minimum intra-cluster distance;
presetting a data quality threshold, and extracting control behavior information if the data quality value is greater than or equal to the data quality threshold And stores the output information setIn the process, otherwise, the control behavior information is removed；
Up to the control action setAfter each piece of control behavior information participates in iteration, the iteration is stopped;
According to the method, in the imaging process, the infrared thermal imager is used, the imaging effect is different due to the change of different positions and different angles of a focal plane on hardware, and the defect on the hardware is optimized through different index parameters on software; the data quality iterative analysis model screens out the hardware control mode of the focal plane with similar characteristics through the similarity analysis of the intra-cluster distance; and then limited by hardware functions, and not satisfying control behaviors with similar characteristics, can exhibit the same imaging effect.
Further, the implementation process of the step S300 includes:
step S301: for output information set Performing behavior classification, using the position information of the same focal plane as a first classification reference, traversing label pairs with the same first classification reference, generating a first classification set, and recording asWherein, the method comprises the steps of, wherein,Representing position information in focal planeCorrespondingly generating a first classification set for the first classification reference;
step S302: traversing tag pairs with the same second classification references by taking the angle information of the same focal plane as the second classification references, and generating a second classification set which is recorded as Wherein, the method comprises the steps of, wherein,Representing angular information in the focal planeFor the second class reference, the generated second class set corresponds.
Further, the specific implementation process of the step S400 includes:
step S401: establishing an evaluation index parameter category information base, wherein index parameter category information for evaluating the control effect is stored in the evaluation index parameter category information base; uniformly coding the evaluation index parameters to form an evaluation index parameter set with unique coding attribute, and recording the evaluation index parameter set as Wherein, the method comprises the steps of, wherein,Represents the a-th evaluation index parameter, A represents the total number of evaluation index parameters;
Step S402: by evaluating sets of index parameters Each of the evaluation index parameter pairs controls behavior informationEvaluating to obtain control behavior informationIs a mean value of the initial evaluation of (a); obtaining position information based on initial evaluation mean value of each control behavior information in the first classification setIs recorded as the comprehensive evaluation mean value of (2); Obtaining angle information based on the initial evaluation mean value of each piece of control behavior information in the second classification setIs recorded as the comprehensive evaluation mean value of (2)；
Respectively byAndCalibrating the positions and angles of the focusing planes of the corresponding position labels and the angle labels;
according to the method, each evaluation index parameter participates in the evaluation of the imaging effect in software, but the evaluation values always have differences due to the limitation of focal plane hardware, and the differences are difficult to optimize through unified index parameters; based on different states of the focal plane, the evaluation values of imaging effects of different index parameters are respectively obtained, and the unified index parameters in the maximum range can be limited based on unified hardware, so that the differentiation of combination regulation of software and hardware is reduced as much as possible; And Different optimal achievement evaluation values are obtained under unified index parameters which respectively form a maximum range by unified position limitation and unified angle limitation; furthermore, the position limitation and the angle limitation are combined, the unification on the hardware limitation is realized, and meanwhile, the self-adaptive optimal regulation and control on the software is realized by combining the unification of the hardware limitation.
Compared with the prior art, the invention has the following beneficial effects: the invention provides a multi-parameter anomaly calibration system and a multi-parameter anomaly calibration method applied to a thermal infrared imager, which are used for generating a position label and an angle label of a focal plane based on the internal structural design of the thermal infrared imager and recording control behavior information of a user side; randomizing the control behavior information into a plurality of groups, wherein each group forms a sub-set data cluster, analyzing the data quality, and carrying out quality screening on the control behavior information; classifying the control behavior information based on the screening result; evaluating the control behavior information through an evaluation index, and calibrating the control behavior based on an evaluation result; therefore, the method can make up the limitation of hardware, realize self-adaptive optimal regulation and control on software, and can realize the maximum range of unified index parameters based on the unified hardware limitation, so that the differentiation of the combination regulation and control of the software and the hardware is reduced as much as possible, and the abnormal calibration of the thermal infrared imager is realized.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention. In the drawings:
FIG. 1 is a schematic diagram of a multi-parameter anomaly calibration system for a thermal infrared imager according to the present invention;
fig. 2 is a schematic step diagram of a multi-parameter anomaly calibration method applied to a thermal infrared imager.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1-2, the present invention provides the following technical solutions:
referring to fig. 1, in a first embodiment: provided is a multi-parameter anomaly calibration system applied to a thermal infrared imager, which comprises: the system comprises an operation log module, a data quality analysis module, a data processing module and a parameter calibration module;
the operation log module is used for generating a position label and an angle label of a focal plane based on the internal structural design of the thermal infrared imager and recording control behavior information of a user side;
the operation log module comprises a cache library unit and a tag pair unit;
The cache library unit is used for establishing an operation log cache library, and the operation log cache library stores control behavior information during each dynamic adjustment of imaging effect, wherein the control behavior information comprises position information of a focal plane and angle information of the focal plane; based on the internal structural design of the thermal infrared imager, respectively establishing a position tag and an angle tag of a focal plane in a unified form, wherein the position tag and the angle tag respectively have unique tag attributes;
a tag pair unit for marking the ith position tag as The j-th angle label is marked asCollecting control behavior information during the kth dynamic adjustment of imaging effect, generating a control behavior set, and recording asAnd (2) andWherein, the method comprises the steps of, wherein,E represents the E-th control behavior information at the kth time of dynamically adjusting the imaging effect, E represents the total number of control behavior information at the kth time of dynamically adjusting the imaging effect,Position tag representing focal planeAngle label for sum focal planeA label pair is formed;
The data quality analysis module is used for randomizing the control behavior information into a plurality of groups, forming a subset data cluster by each group, analyzing the data quality and carrying out quality screening on the control behavior information;
The data quality analysis module comprises a subset forming unit and a quality analysis unit;
The subset forming unit is used for constructing a behavior representation two-dimensional coordinate system based on the control behavior information, wherein the abscissa of the behavior representation two-dimensional coordinate system corresponds to the position label of the focal plane, and the ordinate of the behavior representation two-dimensional coordinate system corresponds to the angle label of the focal plane; marking control behavior information in the behavior representation two-dimensional coordinate system based on the behavior representation two-dimensional coordinate system; for control action set Each control behavior information in the data cluster is randomly grouped to form a subset data cluster, and any subset data cluster is obtained and recorded asAnd (2) and，X represents the number of the subset data clusters, and y represents the total number of the subset data clusters; to control behavior informationAs a subset data clusterRespectively obtain subset data clustersPunctuation corresponding to each control behavior information and control behavior informationEuclidean distance between corresponding punctuations, based on the Euclidean distance, a subset data cluster is obtainedIs recorded as the intra-cluster distance of (2)；
The quality analysis unit is used for constructing a data quality iterative analysis model, and controlling the behavior set during each iterationA sample input is selected at will, and the sample input selected in each iteration is different;
Control behavior information As a sample input for the r-th iteration, a data quality value is calculated, and a specific calculation formula is as follows:
wherein, Representing control behavior informationIs used for the data quality value of (a),A desired value representing a minimum intra-cluster distance;
presetting a data quality threshold, and extracting control behavior information if the data quality value is greater than or equal to the data quality threshold And stores the output information setIn the process, otherwise, the control behavior information is removed；
Up to the control action setAfter each piece of control behavior information participates in iteration, the iteration is stopped;
the data processing module classifies the control behavior information based on the screening result;
the data processing module comprises a first classifier unit and a second classifier unit;
A first classifier unit for outputting information sets Performing first behavior classification, taking the position information of the same focal plane as a first classification reference, traversing label pairs with the same first classification reference, generating a first classification set, and recording asWherein, the method comprises the steps of, wherein,Representing position information in focal planeCorrespondingly generating a first classification set for the first classification reference;
A second classifier unit for outputting information sets Performing second behavior classification, traversing tag pairs with the same second classification references by taking angle information of the same focal plane as the second classification references, generating a second classification set, and recording asWherein, the method comprises the steps of, wherein,Representing angular information in the focal planeCorrespondingly generating a second classification set for the second classification reference;
the parameter calibration module is used for evaluating the control behavior information through the evaluation index and calibrating the control behavior based on the evaluation result
The parameter calibration module comprises an index parameter unit and a calibration analysis unit;
The index parameter unit is used for establishing an evaluation index parameter category information base, and index parameter category information for evaluating the control effect is stored in the evaluation index parameter category information base; uniformly coding the evaluation index parameters to form an evaluation index parameter set with unique coding attribute, and recording the evaluation index parameter set as Wherein, the method comprises the steps of, wherein,Represents the a-th evaluation index parameter, A represents the total number of evaluation index parameters;
calibration analysis unit for evaluating index parameter set Each of the evaluation index parameter pairs controls behavior informationEvaluating to obtain control behavior informationIs a mean value of the initial evaluation of (a); obtaining position information based on initial evaluation mean value of each control behavior information in the first classification setIs recorded as the comprehensive evaluation mean value of (2); Obtaining angle information based on the initial evaluation mean value of each piece of control behavior information in the second classification setIs recorded as the comprehensive evaluation mean value of (2)；
Respectively byAndAnd calibrating the positions and angles of the focusing planes of the corresponding position labels and the angle labels.
Referring to fig. 2, in the second embodiment: the multi-parameter anomaly calibration method applied to the thermal infrared imager comprises the following steps of:
Step S100: based on the internal structural design of the thermal infrared imager, generating a position label and an angle label of a focal plane, and recording control behavior information of a user side;
specifically, an operation log buffer library is established, and control behavior information during each dynamic adjustment of imaging effect is stored in the operation log buffer library, wherein the control behavior information comprises focal plane position information and focal plane angle information; based on the internal structural design of the thermal infrared imager, respectively establishing a position tag and an angle tag of a focal plane in a unified form, wherein the position tag and the angle tag respectively have unique tag attributes;
marking the ith position tag as The j-th angle label is marked asCollecting control behavior information during the kth dynamic adjustment of imaging effect, generating a control behavior set, and recording asAnd (2) andWherein, the method comprises the steps of, wherein,E represents the E-th control behavior information at the kth time of dynamically adjusting the imaging effect, E represents the total number of control behavior information at the kth time of dynamically adjusting the imaging effect,Position tag representing focal planeAngle label for sum focal planeA label pair is formed;
Step S200: randomizing the control behavior information into a plurality of groups, wherein each group forms a sub-set data cluster, analyzing the data quality, and carrying out quality screening on the control behavior information;
Specifically, a behavior representation two-dimensional coordinate system is constructed based on control behavior information, the abscissa of the behavior representation two-dimensional coordinate system corresponds to a position label of a focal plane, and the ordinate of the behavior representation two-dimensional coordinate system corresponds to an angle label of the focal plane; marking control behavior information in the behavior representation two-dimensional coordinate system based on the behavior representation two-dimensional coordinate system; for control action set Each control behavior information in the data cluster is randomly grouped to form a subset data cluster, and any subset data cluster is obtained and recorded asAnd (2) and，X represents the number of the subset data clusters, and y represents the total number of the subset data clusters; to control behavior informationAs a subset data clusterRespectively obtain subset data clustersPunctuation corresponding to each control behavior information and control behavior informationEuclidean distance between corresponding punctuations, based on the Euclidean distance, a subset data cluster is obtainedIs recorded as the intra-cluster distance of (2)；
Constructing a data quality iterative analysis model, and controlling a behavior set during each iterationA sample input is selected at will, and the sample input selected in each iteration is different;
Control behavior information As a sample input for the r-th iteration, a data quality value is calculated, and a specific calculation formula is as follows:
wherein, Representing control behavior informationIs used for the data quality value of (a),A desired value representing a minimum intra-cluster distance;
presetting a data quality threshold, and extracting control behavior information if the data quality value is greater than or equal to the data quality threshold And stores the output information setIn the process, otherwise, the control behavior information is removed；
Up to the control action setAfter each piece of control behavior information participates in iteration, the iteration is stopped;
step S300: classifying the control behavior information based on the screening result;
specifically, for output information set Performing behavior classification, using the position information of the same focal plane as a first classification reference, traversing label pairs with the same first classification reference, generating a first classification set, and recording asWherein, the method comprises the steps of, wherein,Representing position information in focal planeCorrespondingly generating a first classification set for the first classification reference;
Traversing tag pairs with the same second classification references by taking the angle information of the same focal plane as the second classification references, and generating a second classification set which is recorded as Wherein, the method comprises the steps of, wherein,Representing angular information in the focal planeCorrespondingly generating a second classification set for the second classification reference;
step S400: evaluating the control behavior information through an evaluation index, and calibrating the control behavior based on an evaluation result;
Specifically, an evaluation index parameter category information base is established, and index parameter category information for evaluating the control effect is stored in the evaluation index parameter category information base; uniformly coding the evaluation index parameters to form an evaluation index parameter set with unique coding attribute, and recording the evaluation index parameter set as Wherein, the method comprises the steps of, wherein,Represents the a-th evaluation index parameter, A represents the total number of evaluation index parameters;
For example, index categories include sharpness, depth of field, distortion, and aberrations; meanwhile, parameters corresponding to each index category have the characteristic of diversification;
By evaluating sets of index parameters Each of the evaluation index parameter pairs controls behavior informationEvaluating to obtain control behavior informationIs a mean value of the initial evaluation of (a); obtaining position information based on initial evaluation mean value of each control behavior information in the first classification setIs recorded as the comprehensive evaluation mean value of (2); Obtaining angle information based on the initial evaluation mean value of each piece of control behavior information in the second classification setIs recorded as the comprehensive evaluation mean value of (2)；
Respectively byAndCalibrating the positions and angles of the focusing planes of the corresponding position labels and the angle labels;
For example, when the imaging effect is dynamically adjusted for the 5 th time, the imaging effect under e=20 control actions is obtained through the regulation and control of a large number of control actions, the adaptive learning is performed on the basis of the 20 control actions on the algorithm, and finally the imaging effect is obtained AndCorresponding position labelAnd angle labelThe position and angle of the focal plane are calibrated.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Finally, it should be noted that: the foregoing description is only a preferred embodiment of the present invention and is not intended to limit the present invention, but although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the technical solutions described in the foregoing embodiments, or equivalents may be substituted for some of the technical features thereof. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (2)
1. The multi-parameter anomaly calibration method applied to the thermal infrared imager is characterized by comprising the following steps of:
Step S100: based on the internal structural design of the thermal infrared imager, generating a position label and an angle label of a focal plane, and recording control behavior information of a user side;
Step S200: randomizing the control behavior information into a plurality of groups, wherein each group forms a sub-set data cluster, analyzing the data quality, and carrying out quality screening on the control behavior information;
step S300: classifying the control behavior information based on the screening result;
step S400: evaluating the control behavior information through an evaluation index, and calibrating the control behavior based on an evaluation result;
the specific implementation process of the step S100 includes:
Step S101: establishing an operation log cache, wherein the operation log cache stores control behavior information during each dynamic adjustment of imaging effect, and the control behavior information comprises position information of a focal plane and angle information of the focal plane; based on the internal structural design of the thermal infrared imager, respectively establishing a position tag and an angle tag of a focal plane in a unified form, wherein the position tag and the angle tag respectively have unique tag attributes;
Step S102: the ith position label is marked as S i, the jth angle label is marked as U j, control behavior information during the kth dynamic adjustment imaging effect is collected, a control behavior set is generated and marked as CB k＝{tte |e E [1, E ] }, and tt e＝[Si,Uj ], wherein tt e represents the ith control behavior information during the kth dynamic adjustment imaging effect, E represents the total quantity of the control behavior information during the kth dynamic adjustment imaging effect, and [ S i,Uj ] represents a label pair formed by a focal plane position label S i and a focal plane angle label U j;
the specific implementation process of the step S200 includes:
Step S201: constructing a behavior representation two-dimensional coordinate system based on control behavior information, wherein the abscissa of the behavior representation two-dimensional coordinate system corresponds to a position label of a focal plane, and the ordinate of the behavior representation two-dimensional coordinate system corresponds to an angle label of the focal plane; marking control behavior information in the behavior representation two-dimensional coordinate system based on the behavior representation two-dimensional coordinate system; randomly grouping the control behavior information in the control behavior set CB k to form a subset data cluster, obtaining any subset data cluster, marking as CB x, and X represents the number of the subset data clusters, and y represents the total number of the subset data clusters; taking the control behavior information tt e as the centroid of the subset data cluster cb x, respectively obtaining the Euclidean distance between the punctuation corresponding to each control behavior information and the punctuation corresponding to the control behavior information tt e in the subset data cluster cbx, and obtaining the intra-cluster distance of the subset data cluster cb x based on the Euclidean distance, and marking as D (cb x|tte);
Step S202: constructing a data quality iterative analysis model, wherein a sample input is arbitrarily selected from a control behavior set CB k during each iteration, and the sample inputs selected during each iteration are different;
the control behavior information tt e is taken as a sample input of the r-th iteration, and the data quality value is calculated according to the following specific calculation formula:
Wherein DQ (tt e) represents a data quality value of the control behavior information tt e, DQ represents an expected value of the minimum intra-cluster distance;
Presetting a data quality threshold, extracting control behavior information tt e and storing the control behavior information tt e into an output information set OUT k if the data quality value is larger than or equal to the data quality threshold, otherwise eliminating the control behavior information tt e;
until each piece of control behavior information in the control behavior set CB k participates in iteration, stopping the iteration;
the specific implementation process of the step S300 includes:
Step S301: performing behavior classification on the output information set OUT k, traversing tag pairs with the same first classification references by taking the position information of the same focal plane as the first classification references, and generating a first classification set which is marked as FTS (S i), wherein FTS (S i) represents the first classification references by taking the position information S i of the focal plane as the first classification references, and correspondingly generating the first classification set;
Step S302: traversing tag pairs with the same second classification references by taking the angle information of the same focal plane as the second classification references, and generating a second classification set which is marked as STS (U j), wherein STS (U j) represents the second classification set which is correspondingly generated by taking the angle information U j of the focal plane as the second classification reference;
The specific implementation process of the step S400 includes:
Step S401: establishing an evaluation index parameter category information base, wherein index parameter category information for evaluating the control effect is stored in the evaluation index parameter category information base; uniformly encoding the evaluation index parameters to form an evaluation index parameter set with unique encoding attributes, wherein the evaluation index parameter set is marked as IC= { IC a |a epsilon [1, A ] }, wherein IC a represents the a-th evaluation index parameter, and A represents the total number of the evaluation index parameters;
Step S402: evaluating the control behavior information tt e through each evaluation index parameter in the evaluation index parameter set IC to obtain an initial evaluation mean value of the control behavior information tt e; based on the initial evaluation mean value of each piece of control behavior information in the first classification set, the comprehensive evaluation mean value of the obtained position information S i is recorded as P 1(Si); based on the initial evaluation mean value of each piece of control behavior information in the second classification set, obtaining a comprehensive evaluation mean value of angle information U j, and recording as P 2(Uj);
The positions and angles of the focusing planes of the position labels and the angle labels corresponding to max { P 1(Si) } and max { P 2(Uj) } are calibrated respectively.
2. A multi-parameter anomaly calibration system for a thermal infrared imager, the system comprising: the system comprises an operation log module, a data quality analysis module, a data processing module and a parameter calibration module;
The operation log module is used for generating a position label and an angle label of a focal plane based on the internal structural design of the thermal infrared imager and recording control behavior information of a user side;
the data quality analysis module is used for randomizing the control behavior information into a plurality of groups, forming a subset data cluster by each group, analyzing the data quality and carrying out quality screening on the control behavior information;
The data processing module classifies the control behavior information based on the screening result;
the parameter calibration module is used for evaluating the control behavior information through the evaluation index and calibrating the control behavior based on the evaluation result;
The operation log module comprises a cache library unit and a tag pair unit;
The cache library unit is used for establishing an operation log cache library, wherein the operation log cache library stores control behavior information during each dynamic adjustment of imaging effect, and the control behavior information comprises position information of a focal plane and angle information of the focal plane; based on the internal structural design of the thermal infrared imager, respectively establishing a position tag and an angle tag of a focal plane in a unified form, wherein the position tag and the angle tag respectively have unique tag attributes;
The label pair unit is configured to record an ith position label as S i, record a jth angle label as U j, collect control behavior information during kth dynamic adjustment of imaging effects, generate a control behavior set, and record CB k＝{tte |e E [1, E ] }, and tt e＝[Si,Uj, where tt e represents the ith control behavior information during kth dynamic adjustment of imaging effects, E represents the total amount of control behavior information during kth dynamic adjustment of imaging effects, [ S i,Uj ] represents a label pair formed by a focal plane position label S i and a focal plane angle label U j;
the data quality analysis module comprises a subset forming unit and a quality analysis unit;
The subset forming unit is used for constructing a behavior representation two-dimensional coordinate system based on control behavior information, wherein the abscissa of the behavior representation two-dimensional coordinate system corresponds to a position label of a focal plane, and the ordinate of the behavior representation two-dimensional coordinate system corresponds to an angle label of the focal plane; marking control behavior information in the behavior representation two-dimensional coordinate system based on the behavior representation two-dimensional coordinate system; randomly grouping the control behavior information in the control behavior set CB k to form a subset data cluster, obtaining any subset data cluster, marking as CB x, and X represents the number of the subset data clusters, and y represents the total number of the subset data clusters; taking the control behavior information tt e as the centroid of the subset data cluster cb x, respectively obtaining the Euclidean distance between the punctuation corresponding to each control behavior information in the subset data cluster cb x and the punctuation corresponding to the control behavior information tt e, and obtaining the intra-cluster distance of the subset data cluster cb x based on the Euclidean distance, and marking as D (cb x|tte);
The quality analysis unit is used for constructing a data quality iterative analysis model, wherein one sample input is arbitrarily selected from the control behavior set CB k during each iteration, and the sample inputs selected during each iteration are different;
the control behavior information tt e is taken as a sample input of the r-th iteration, and the data quality value is calculated according to the following specific calculation formula:
Wherein DQ (tt e) represents a data quality value of the control behavior information tt e, DQ represents an expected value of the minimum intra-cluster distance;
Presetting a data quality threshold, extracting control behavior information tt e and storing the control behavior information tt e into an output information set OUT k if the data quality value is larger than or equal to the data quality threshold, otherwise eliminating the control behavior information tt e;
until each piece of control behavior information in the control behavior set CB k participates in iteration, stopping the iteration;
The data processing module comprises a first classifier unit and a second classifier unit;
The first classifier unit is configured to perform a first behavior classification on the output information set OUT k, traverse tag pairs with the same first classification reference by using the position information of the same focal plane as the first classification reference, and generate a first classification set, and record the first classification set as FTS (S i), where FTS (S i) represents the first classification reference by using the position information S i of the focal plane, and corresponds to the generated first classification set;
the second classifier unit is configured to classify the output information set OUT k according to a second behavior, take the angle information of the same focal plane as a second classification reference, traverse tag pairs with the same second classification reference, and generate a second classification set, and record the second classification set as STS (U j), where STS (U j) represents the second classification set correspondingly generated by taking the angle information U j of the focal plane as the second classification reference;
The parameter calibration module comprises an index parameter unit and a calibration analysis unit;
The index parameter unit is used for establishing an evaluation index parameter category information base, and index parameter category information for evaluating the control effect is stored in the evaluation index parameter category information base; uniformly encoding the evaluation index parameters to form an evaluation index parameter set with unique encoding attributes, wherein the evaluation index parameter set is marked as IC= { IC a |a epsilon [1, A ] }, wherein IC a represents the a-th evaluation index parameter, and A represents the total number of the evaluation index parameters;
The calibration analysis unit is configured to evaluate the control behavior information tt e through each evaluation index parameter in the evaluation index parameter set IC, so as to obtain an initial evaluation mean value of the control behavior information tt e; based on the initial evaluation mean value of each piece of control behavior information in the first classification set, the comprehensive evaluation mean value of the obtained position information S i is recorded as P 1(Si); based on the initial evaluation mean value of each piece of control behavior information in the second classification set, obtaining a comprehensive evaluation mean value of angle information U j, and recording as P 2(Uj);
The positions and angles of the focusing planes of the position labels and the angle labels corresponding to max { P 1(Si) } and max { P 2(Uj) } are calibrated respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410552357.4A CN118225254B (en) | 2024-05-07 | 2024-05-07 | Multi-parameter anomaly calibration system and method applied to thermal infrared imager |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410552357.4A CN118225254B (en) | 2024-05-07 | 2024-05-07 | Multi-parameter anomaly calibration system and method applied to thermal infrared imager |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118225254A CN118225254A (en) | 2024-06-21 |
| CN118225254B true CN118225254B (en) | 2024-11-08 |
Family
ID=91511972
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410552357.4A Active CN118225254B (en) | 2024-05-07 | 2024-05-07 | Multi-parameter anomaly calibration system and method applied to thermal infrared imager |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118225254B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116305705A (en) * | 2021-12-20 | 2023-06-23 | 中国科学院沈阳自动化研究所 | Imaging system environment adaptability assessment method based on infrared simulation |
| CN116720118A (en) * | 2023-06-13 | 2023-09-08 | 平安科技（深圳）有限公司 | Label quality intelligent analysis method and device, electronic equipment and storage medium |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102066157B1 (en) * | 2018-06-12 | 2020-01-14 | 중앙대학교 산학협력단 | Method and apparatus for generating feature subset in classifying multi-label pattern |
| WO2022113219A1 (en) * | 2020-11-25 | 2022-06-02 | 株式会社日立製作所 | Tag domain presentation device and tag domain presentation method, and information processing system using same |
| CN117323578B (en) * | 2023-12-01 | 2024-01-30 | 深圳恩多克医疗有限公司 | Intelligent control method and system for infrared therapeutic apparatus |
| CN117563144B (en) * | 2023-12-04 | 2024-05-28 | 郭永强 | Method and system for evaluating condition and predicting residual life of infrared therapeutic instrument |
-
2024
- 2024-05-07 CN CN202410552357.4A patent/CN118225254B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116305705A (en) * | 2021-12-20 | 2023-06-23 | 中国科学院沈阳自动化研究所 | Imaging system environment adaptability assessment method based on infrared simulation |
| CN116720118A (en) * | 2023-06-13 | 2023-09-08 | 平安科技（深圳）有限公司 | Label quality intelligent analysis method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118225254A (en) | 2024-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109190665B (en) | Universal image classification method and device based on semi-supervised generation countermeasure network | |
| CN111126386B (en) | Sequence domain adaptation method based on countermeasure learning in scene text recognition | |
| CN109783879B (en) | Radar radiation source signal identification efficiency evaluation method and system | |
| KR20220095216A (en) | BBP-assisted defect detection flow for SEM images | |
| CN109726195B (en) | Data enhancement method and device | |
| CN107679550A (en) | A kind of appraisal procedure of data set classification availability | |
| CN112819063B (en) | Image identification method based on improved Focal loss function | |
| Zhang | Application of artificial intelligence recognition technology in digital image processing | |
| CN118225254B (en) | Multi-parameter anomaly calibration system and method applied to thermal infrared imager | |
| CN110956277A (en) | Interactive iterative modeling system and method | |
| CN116721343A (en) | Cross-domain field cotton boll recognition method based on deep convolutional neural network | |
| CN113377991B (en) | Image retrieval method based on most difficult positive and negative samples | |
| CN110471854A (en) | A kind of defect report assigning method based on high dimensional data mixing reduction | |
| CN111031042A (en) | Network anomaly detection method based on improved D-S evidence theory | |
| CN116560575A (en) | Financial data processing system based on big data and cloud computing | |
| CN115953245A (en) | Stock trend prediction method and device based on sequence-to-graph | |
| CN109359694B (en) | Image classification method and device based on mixed collaborative representation classifier | |
| Roy et al. | New crossover operators for multiple subset selection tasks | |
| CN113656480A (en) | Data integration method and system for heterogeneous data source | |
| CN112613521A (en) | Multilevel data analysis system and method based on data conversion | |
| Chen et al. | An assessment of qualitative performance of machine learning architectures: Modular feedback networks | |
| CN117746266B (en) | Tree crown detection method, device and medium based on semi-supervised interactive learning | |
| CN111882061B (en) | Convolutional neural network training method based on hierarchical random gradient descent | |
| JP7322918B2 (en) | Program, information processing device, and learning model generation method | |
| US20230186047A1 (en) | Evaluation method, evaluation device, and computer program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |